Technical Field
The present invention relates to AC-DC converters.
Background Art
FIG. 10 is a circuit diagram of a conventional AC-DC converter, a circuit well known as a PFC (Power Factor Correction) circuit.
In FIG. 10, 1 is an AC power source; 2 to 5 are diodes constituting a rectifier circuit (diode bridge) DB; 6 is an inductor; 7 is a semiconductor switching device; 8 is a diode; 9 is a capacitor; and 10 is a load. The portion constituted by the inductor 6, the switching device 7, the diode 8, and the capacitor 9 is also known as a so-called boost converter, which boosts a DC voltage to increase output. Other than the MOSFET (Metal Oxide Silicon Field Effect Transistor) shown here in the diagram, an IGBT (Isolated Gate Bipolar Transistor), a BJT (Bipolar Junction Transistor), or the like may also be used for the switching device 7.
Note that in FIG. 10, Vin is an AC input voltage; Iin is an AC input current; IL is a current flowing through the inductor 6; Vr1 is an output voltage of the rectifier circuit DB; Vr2 is a drain-source voltage of the switching device 7; and E is a DC output voltage (terminal voltage across the capacitor 9).
Functions of the circuit shown in FIG. 10 are as follows:                (1) To convert the AC input voltage Vin into a DC output voltage E of a desired voltage and to keep the DC output voltage E constant regardless of fluctuations in the AC input voltage Vin and load current.        (2) To make the AC input current Iin a sinusoidal wave with a power factor of approximately 1.        
Operations for achieving the aforementioned functions (1) and (2) will be described with reference to FIGS. 10 to 12. Note that the forward voltage drops of the diodes 2 to 5 and 8 and the switching device 7 will be disregarded in the descriptions below.
If the AC input voltage Vin at present is assumed to be a sinusoidal wave such as the one shown in FIG. 12, the voltage Vr1 outputted by the diode bridge DB becomes a fully rectified waveform.
When the AC input voltage Vin has positive polarity and the switching device 7 in FIG. 10 is turned on, the voltage Vr2 becomes 0V and the current Iin flows in the following path: AC power source 1→diode 2→inductor 6→switching device 7→diode 5→AC power source 1. As a result of this, the voltage Vin is applied to both ends of the inductor 6, and the current IL increases. Meanwhile, when the switching device 7 is turned off, the current Iin flows in the following path: AC power source 1→diode 2→inductor 6→diode 8→capacitor 9→diode 5→AC power source 1. At this time, the voltage Vr2 is nearly equal to the terminal voltage E across the capacitor 9, and a voltage that is a difference between the voltage E and the AC input voltage Vin is applied to the inductor 6. Note the circuit operates so as to keep the voltage E higher than the peak value of the AC input voltage Vin. As a result, the IL decreases.
It follows that, by controlling the on-off time ratio of the switching device 7, it is possible to control the waveform and the size of the current IL in any manner. If the current IL is set to be a rectified sinusoidal waveform similar to that of the voltage Vr1 (here, the ripple is disregarded), the AC input current Iin becomes a sinusoidal waveform. Additionally, by controlling the amplitude of the current IL according to load power, it is possible to maintain the DC output voltage E to a desired constant value.
FIG. 11 is a block diagram of a control circuit for controlling the on-off time ratio of the switching device 7. In FIGS. 11, 102 to 105 are adders; 106 is an absolute value calculator; 107 is a voltage regulator (AVR); 108 is a multiplier; 109 is a current regulator (ACR); 111 is a comparator; 112 is a logic inverter; and 113 is a carrier generator that produces triangle waves.
The control circuit operates as follows:
The control circuit detects the DC output voltage E in FIG. 10 using a known voltage detector, obtains a deviation of the voltage E from a command value E* using the adder 102, and inputs the obtained deviation into the voltage regulator 107. The voltage regulator 107 causes an amplitude command of the current IL to increase if the voltage E is less than the command value E*, and causes the amplitude command of the current IL to decrease when the voltage E is more than the command value E*. A PI (proportional-integral) controller is used for the voltage regulator 107, for example.
Meanwhile, the control circuit detects the AC input voltage Vin using a known voltage detector, and obtains the absolute value using the absolute value calculator 106. An output of the absolute value calculator 106 is a waveform roughly similar to that of the voltage Vr1, if the forward voltage drops of the diodes 2 to 5 are disregarded. By multiplying the output of the absolute value calculator 106 to the amplitude command of the current IL using the multiplier 108, an instantaneous value command of the current IL is obtained.
Further, the control circuit detects the current IL using a known current detector, calculates the deviation of the current IL from the aforementioned instantaneous value command using the adder 103, and inputs the obtained deviation into the current regulator 109. The current regulator 109 causes output to increase if the current IL is less than the instantaneous value command, and causes output to decrease if the IL is more than the instantaneous value command. A P (proportional) regulator is used as the current regulator 109, for example.
Next, an instantaneous value command of the voltage Vr2 is obtained by adding the absolute value of the AC input voltage Vin and the output of the current regulator 109 using the adder 104. Here, the sign of the output of the current regulator 109 is inverted before the output is added to the absolute value of the voltage V. For this reason, the instantaneous value command of the voltage Vr2 decreases when the output of the current regulator 109 increases as a result of an insufficient current IL, thereby enlarging the difference between the Vr1 and the Vr2. This results in a larger current flowing into the side of the switching device 7.
Further, by treating the instantaneous value command of the voltage Vr2 as a signal wave, and by comparing the signal wave with a triangular wave carrier outputted by the carrier generator 113 using the adder 105 and the comparator 111, a PWM (Pulse Width Modulation) is performed. Then the output of the comparator 111 is inputted into the logic inverter 112 to generate a gate signal for the switching device 7.
In other words, as shown in FIG. 12, when the signal wave (instantaneous value command of Vr2)>carrier, the switching device 7 is turned off and Vr2 (PWM pulse)=E, and when the signal wave<carrier, the switching device 7 is turned on and Vr2=0V. As a result, the voltage Vr2 of FIG. 10 becomes a series of pulses such as the one shown in FIG. 12. The low-frequency components of Vr2, excluding the switching frequency component, are similar to those of the voltage Vr1, and have a waveform that is slightly out of phase from the waveform of Vr1. When a difference in voltage resulting from this phase difference is applied to the inductor 6 in FIG. 10, the current IL flows, and the current IL as a result has a waveform similar to that of the voltage Vr1.
In FIG. 10, when the AC input voltage Vin has positive polarity, the diodes 2 and 5 are electrically connected and the AC input current Iin and the IL have identical polarity. When the voltage Vin has negative polarity, the diodes 3 and 4 are electrically connected and the currents Iin and IL have opposite polarities. As a result, the AC input current Iin has a sinusoidal wave with a power factor of 1 with a phase that is substantially identical to that of the AC input voltage Vin.
Additionally, by the control operation described above, when the DC output voltage E is insufficient, the amplitude of the current Iin is increased, and a larger amount of power flows into the circuit from the AC power source 1, thereby resulting in an increased value of the voltage E. As a result, the voltage E is kept at a desired constant value.
Note that, in FIG. 12, the carrier frequency, or the switching frequency, is indicated as approximately a few multiples of the frequency of the AC input voltage Vin so that it is easier to view the diagram. However, in an actual device, it is common to make the carrier frequency at least 100 times the frequency of the AC input voltage Vin (for example, if the frequency of Vin is 50 Hz, the carrier frequency is 5 kHz or higher), so that the ripple current is kept sufficiently small even if the inductor 6 is downsized.
Meanwhile, the withstand voltages of semiconductor devices such as the diodes 2 to 5 and 8 and the switching device 7 shown in FIG. 10 need to be at least greater than the voltage E, while the voltage E must be greater than the peak value of the AC input voltage Vin. In a circuit with an AC input voltage Vin of 200V or less (effective value; hereinafter the same), for example, semiconductor devices with a withstand voltage that is greater than the voltage E (which exceeds the peak value of the Vin) and is no more than 600V are normally used. In a circuit with an AC input voltage Vin of 400V or less, semiconductor devices with a withstand voltage that is greater than the voltage E (which exceeds the peak value of the Vin) and is no more than 1,200V are normally used.
For this reason, when the AC input voltage Vin exceeds 400V, semiconductors with a withstand voltage exceeding 1,200V will be required.
However, there is significant switching loss for semiconductor devices with a withstand voltage exceeding 1,200V, particularly for the switching devices 7 and the diode 8 operating at high frequency, resulting in a decreased efficiency of the device as a whole. Additionally, only a few types of semiconductor devices with such a high withstand voltage are available commercially, and it is difficult to select a semiconductor device that is appropriate in design while considering the current rating, outer geometry, price, and the like.
As a solution to this problem, there is a technology that achieves a desired withstand voltage by connecting a plurality of semiconductor devices in series. However, there is a risk of applying an overvoltage to some of the semiconductor devices unless the switching timing of the semiconductor devices are synched accurately. For this reason, applying this solution to a device that performs high-frequency switching will be difficult.
The circuit shown in FIG. 13 is a known example of a conventional technology for avoiding an increase in the withstand voltage of a semiconductor device such as the one described above.
In FIGS. 13, 20 and 21 are semiconductor switching devices; 22 and 23 are diodes, and 24 and 25 are capacitors. All other parts are given reference characters identical to those of FIG. 10.
The part constituted by the switching devices 20 and 21, the diodes 22 and 23, and the capacitors 24 and 25 is a known configuration shown in FIG. 1 of Patent Document 1, and is called a 3-level converter.
A DC output voltage E is divided by ½ into voltages Ep and an En by the capacitors 24 and 25. A voltage Vr2 becomes 0V when both of the switching devices 20 and 21 are on, and becomes Ep or En, or E/2, when one of the switching devices 20 and 21 is off. When both of the switching devices 20 and 21 are off, the voltage Vr2 becomes (Ep+En), or E.
Thus, the configuration is called “3-level converter” since there are three levels of the voltage Vr2, 0V, E/2, and E, to choose from.
According to the circuit in FIG. 13, the voltage applied respectively to the series circuit of the switching device 20 and the diode 22 and the series circuit of the switching device 21 and the diode 23 is E/2. Therefore, for identical voltages Vin and E, the withstand voltages of semiconductor devices can be reduced by ½. In other words, if the withstand voltages of the semiconductor devices are identical, the voltages Vin and E can be doubled. Therefore, it is possible to configure a circuit with an AC input voltage Vin of an 800V class using semiconductor devices with a withstand voltage of 1,200V, for example.